CN101195072A - Measuring method and measuring device for motor behavior - Google Patents

Abstract

A measuring device comprises cameras for imaging a person performing a movement from at least two different directions so as to provide image data. An image-processing unit measures the image data for values of predetermined parameters representing motion so as to provide measurement data. The image-processing unit presents the performance in relation to other performances.

Description

Translated fromChinese

运动行为的测量方法和测量装置Measuring method and measuring device for motor behavior

技术领域technical field

本发明涉及用于测量包括运动在内的行为的方法和测量装置。The invention relates to a method and a measuring device for measuring behavior including movement.

背景技术Background technique

例如，为了改善体育运动行为、学习正确的工作人机工程学或为了残疾人的恢复，人们可以测量并分析人体的运动和姿势。例如在体育运动行为过程中，教练员可能只用眼睛观察受训人员的行为并告诉他/她如何改善或提高其行为和/或告诉他或她的行为有哪些不足。For example, one can measure and analyze the movements and postures of the human body in order to improve sports performance, learn correct ergonomics at work or for the rehabilitation of disabled people. For example, during sports performance, a coach may just visually observe a trainee's behavior and tell him/her how to improve or enhance his or her behavior and/or tell him or her where his or her behavior is deficient.

教练员也可能求助于各种不同的测量装置，借此例如可以视频记录下行为。测量装置可以分析图像数据，从行为中确定各种不同的简单特征，教练员可以通过上述简单特征并依据自身的见识尝试指导受训人员来改善其行为。为了成像，可以在四肢、躯体和或许体育运动器材上安装许多标记，这些标记有助于运动的测量。The trainer can also resort to various measuring devices, by means of which, for example, the behavior can be video-recorded. The image data can be analyzed by the measuring device to determine various simple features from the behavior that the trainer can use to try to guide the trainees to improve their behavior based on their own insights. For imaging, a number of markers can be mounted on the extremities, torso, and perhaps sports equipment, which aid in the measurement of motion.

发明内容Contents of the invention

本发明的目的是提供一种改进方法和一种实施该方法的测量装置。通过用于测量包括运动在内的行为的方法来实现上述目的，该方法包括：从至少两个不同的方向对做运动的每个人进行成像，由此提供每个行为的图像数据；为了代表每个运动的预定参数值，测量该图像数据以提供测量数据；关于在先测量的多个行为，显示依据该测量数据的至少一个行为。The object of the invention is to provide an improved method and a measuring device for carrying out the method. The above objects are achieved by a method for measuring behavior including movement, the method comprising: imaging each individual performing movement from at least two different directions, thereby providing image data for each behavior; The image data is measured to provide measurement data for a predetermined parameter value of a movement, and at least one behavior according to the measurement data is displayed with respect to a plurality of previously measured behaviors.

本发明也涉及一种用于测量包括运动在内的行为的测量装置。该测量装置包括：至少两个摄像器材，用于从至少两个不同的方向对做运动的每个人进行成像，由此提供行为的图像数据；图像处理单元，其被构造成从每个图像数据中测量代表运动的预定参数值以提供测量数据，并且关于在先测量的多个行为显示依据该测量数据的至少一个行为。The invention also relates to a measuring device for measuring behavior including movement. The measuring device comprises: at least two camera equipment, for imaging each person doing exercise from at least two different directions, thereby providing image data of behavior; A predetermined parameter value representative of the movement is measured to provide measurement data, and at least one behavior dependent on the measurement data is displayed with respect to a plurality of previously measured behaviors.

通过本发明的方法和装置获得多个优点。可以从图像数据中抽取参数值，对此不用将标记安置到做运动的人上。不需要教练员来分析测量结果或指导如何改善行为，但做运动的人将在每个行为后马上获得易懂的反馈信息。反馈信息是客观的，它只依据测量结果，不依赖教练员的主观意见。Several advantages are obtained by the method and apparatus of the present invention. The parameter values can be extracted from the image data without having to attach markers to the person performing the exercise. There is no need for a trainer to analyze measurements or guide how to improve behavior, but the person doing the exercise will get comprehensible feedback immediately after each behavior. The feedback information is objective, it is only based on the measurement results, and does not rely on the subjective opinions of the coaches.

本发明的方法还包括：将每个测量数据的测量点加到用作行为映射的自组织映射上，并且关于代表该自组织映射上的行为的至少另一个预定神经元显示所述测量点。此外，本发明的方法包括对做运动的人的成像是在没有将图像分析辅助标记安置在人上的情况下进行的。上述方法包括将与每个行为相关联的测量数据传输给数据库以便存储。上述方法可以包括：在存储于该数据库中的测量数据的数据库中产生行为映射并且将该行为映射传输给每个行为地点以便作为基准。本发明的方法还包括：在与每个行为相关联的测量数据的数据库中产生神经元并且由神经元产生用作行为映射的自组织映射。在这里，可以利用因特网将与每个行为相关联的测量数据从各测量点传输给数据库。The method of the invention also includes adding a measurement point of each measurement data to the self-organizing map used as a behavioral map, and displaying said measurement point with respect to at least one other predetermined neuron representing behavior on the self-organizing map. In addition, the method of the present invention includes imaging the person in motion without placing image analysis auxiliary markers on the person. The method described above includes transmitting measurement data associated with each activity to a database for storage. The above method may include generating a behavior map in the database of measurement data stored in the database and transmitting the behavior map to each behavior site for reference. The method of the present invention also includes generating neurons in the database of measurement data associated with each behavior and generating a self-organizing map from the neurons for use as a behavior map. Here, the measurement data associated with each activity can be transmitted from the measurement points to the database using the Internet.

本发明的方法包括：关于在先测量的多个行为，至少在一台计算机上附加显示依据所述测量数据的至少一个行为，其中该计算机通过数据网络与所述图像处理单元通讯。此时，将所述图像处理单元可以被构造成借助计算机来适用于包括运动在内的某个行为。本发明的方法还包括显示对行为的口头意见，该意见依赖于所述测量点同至少另一个行为的关系。The method according to the invention comprises the additional display of at least one behavior based on the measurement data on at least one computer with respect to previously measured behaviors, wherein the computer communicates with the image processing unit via a data network. In this case, the image processing unit can be configured by means of a computer to be suitable for certain activities including sports. The method of the invention also includes displaying a verbal opinion about the behavior, the opinion being dependent on the relationship of said measurement point to at least one other behavior.

在本发明的测量装置中，图像处理单元按以下方式构造，将每个测量数据的测量点加到用作行为映射的自组织映射上，在这里，图像处理单元按以下方式构造，使不同的行为相互成比例并且关于代表自组织映射上的行为的至少另一个预定神经元显示所述测量点。根据本发明，测量装置按以下方式构造，拍摄做运动的人的图像，但不需要将图像分析辅助标记安置在人上。此外，测量装置可以包括数据库，每个行为地点的图像处理单元被构造成将与每个行为相关的测量数据传输给所述数据库以便存储。在这里，数据库按以下方式构造，由存储于数据库中的测量数据产生行为映射并且能传输所述行为映射至每个行为地点以便作为基准。数据库按以下方式构造，由与每个行为相关的测量数据形成神经元并且能由神经元产生用作行为映射的自组织映射。每个图像处理单元按照以下方式构造，用因特网将与每个行为相关的测量数据传输给所述数据库。所述测量装置包括至少一台计算机，所述计算机按以下方式构造，通过数据网络与图像处理单元通讯以显示包括运动在内的行为。此时，图像处理单元可被构造成通过计算机适用于包括运动在内的行为。此外，图像处理单元按以下方式构造，即显示对行为的口头意见，该口头意见依赖于所述测量点同至少另一个行为的关系。In the measuring device of the present invention, the image processing unit is structured in such a way that the measurement point of each measurement data is added to the self-organizing map used as the behavior map, and here, the image processing unit is structured in such a way that different The behaviors are proportional to each other and the measurement point is displayed with respect to at least one other predetermined neuron representing the behavior on the self-organizing map. According to the invention, the measuring device is designed in such a way that an image of a moving person is recorded without the need for image analysis aids to be attached to the person. Furthermore, the measuring device can comprise a database, to which the image processing unit of each activity site is designed to transmit the measurement data relating to each activity for storage. Here, the database is structured in such a way that a behavior map is generated from the measurement data stored in the database and can be transferred to each behavior site for reference. The database is structured in such a way that a neuron is formed from the measurement data associated with each behavior and a self-organizing map that can be used as a behavior map can be generated from the neurons. Each image processing unit is constructed in such a way that the measured data relating to each action are transmitted to the database via the Internet. The measuring device comprises at least one computer configured in such a way that it communicates with an image processing unit via a data network to display behavior including movement. At this time, the image processing unit may be configured to be suitable for actions including motion by a computer. Furthermore, the image processing unit is designed in such a way that it displays a verbal opinion about the behavior which depends on the relationship of the measuring point to at least one other behavior.

附图说明Description of drawings

以下，将结合优选实施例并且参见附图来更详细地说明本发明，其中：Hereinafter, the present invention will be described in more detail in conjunction with preferred embodiments and with reference to the accompanying drawings, wherein:

图1表示测量布置结构；Figure 1 shows the measurement arrangement;

图2表示自组织映射；Figure 2 represents a self-organizing map;

图3A表示测量数据收集；Figure 3A represents measurement data collection;

图3B表示测量装置；Figure 3B shows the measuring device;

图4表示方法的流程图。Figure 4 shows a flow chart of the method.

具体实施方式Detailed ways

图1表示测量装置，它可以被安置在户内或户外的预期地点上。该测量装置包括至少两个摄像器材100、102、104和106，这些摄像器材可以是摄像机并且可以自由定位以便拍摄包括运动在内的行为。例如，摄像器材100、102和104可以按照能拍摄水平方向上的行为的方式定位，而至少一个摄像器材，在这种情况下是摄像器材106可以拍摄竖直方向上的行为。成像也可以斜向下的方式发生。Figure 1 shows the measuring device, which can be placed indoors or outdoors at the desired location. The measuring device comprises at least twocamera devices 100 , 102 , 104 and 106 , which can be video cameras and can be freely positioned in order to record behavior including movement. For example,camera fixtures 100, 102, and 104 may be positioned in such a way as to capture activity in a horizontal direction, while at least one camera fixture, in thiscase camera fixture 106, may capture activity in a vertical direction. Imaging can also take place in a diagonally downward fashion.

在图1中，包括运动在内的行为例如是高尔夫击球，但更宽泛地讲，包括运动在内的行为也可以是在某些其它体育运动中的击打、射击、投掷、跳跃、跨步、踢等。作为可能测量打击的其它体育运动的例子，可以想到拳击和其它体育运动，其训练涉及到用手击打对手，可以想到冰球和其它体育运动，其训练涉及用杆或球拍击打运动器材(例如球或冰球)，还可以想到可能要训练器具投掷的标枪和其它体育运动(如掷链球、推铅球、投镖、棒球、排球、篮球等)、可能要训练跳跃的跳高和其它跳跃体育运动(例如撑杆跳、各种跳水、体操跳跃等)、其训练涉及跨步锻炼的跨栏和其它竞技运动(例如障碍赛)、以及可能要训练有效踢击的足球和其它这样的体育运动(例如空手道和跆拳道等)。In Figure 1, the action that includes the sport is, for example, a golf shot, but more broadly, the action that includes the sport can also be hitting, shooting, throwing, jumping, straddling in some other sport. step, kick, etc. As examples of other sports in which blows may be measured, one can think of boxing and other sports whose training involves hitting an opponent with the hand, ice hockey and other sports whose training involves hitting a sporting device with a pole or racket (e.g. ball or ice hockey), but also conceivable javelin and other sports (such as hammer throw, shot put, dart throwing, baseball, volleyball, basketball, etc.) that may train the throwing of the appliance, high jump and other jumping sports that may train jumping ( such as pole vaulting, various types of diving, gymnastics jumping, etc.), hurdles and other competitive sports in which training involves striding exercises (e.g., obstacle races), and football and other such sports in which effective kicks may be trained (e.g., karate and Taekwondo, etc.).

例如，如果是测量和分析高尔夫击球，则测量装置可以竖立在高尔夫球场的预期地点上。具有所有所需构造的测量装置可以是固定不动的或是活动的，它允许反复的拆装。因此，摄像器材100、102、104和106可以安装在行为地点108的各个侧面的导轨上(图1未示出)，该行为地点就是摄像器材100、102、104和106的拍摄目标。例如，行为地点108可以是例如高为2米至3米且半径为1米至2米的圆锥形空间。因此，做运动的人不需要在行为过程中精确位于某个点上。行为地点108也可以具有不同的形状和尺寸。测量装置也可以包括图像处理单元110，包含来自摄像器材100、102、104和106的图像数据的信号将被输入图像处理单元。For example, if golf shots are being measured and analyzed, the measuring device may be erected at the desired location on the golf course. The measuring device, with all required configurations, can be stationary or movable, allowing repeated disassembly and assembly. Therefore,camera equipment 100, 102, 104 and 106 can be installed on the guide rails (not shown in FIG. For example, theactivity site 108 may be a conical space with a height of 2 meters to 3 meters and a radius of 1 meter to 2 meters. Thus, the person doing the exercise does not need to be at a precise point during the action.Activity venues 108 may also have different shapes and sizes. The measurement device may also include animage processing unit 110 into which signals comprising image data from thecamera equipment 100 , 102 , 104 and 106 are input.

测量地点可以包括墙112和至少一个灯具114、116和118，作为测量装置的附属器材，尽管这些器材不是必需的。为了提供良好的照明和高质量的图像数据，灯具114、116和118可以照亮墙112，来自墙的光和/或其它光学射线在运动中的人120上漫散射。墙112可以包括开口122，人可以经该开口进出测量地。或许存在的运动器材也可能因所做的运动而经开口122离开测量地。除了墙112外，测量地也可以包括顶壁(图1未示出)、为改变亮度而影响照明条件的天篷124、以及用于根据需要关闭开口122的至少一扇门126和128。The measurement site may include awall 112 and at least onelight fixture 114, 116, and 118 as ancillary fixtures for the measurement device, although such fixtures are not required. To provide good lighting and high quality image data,light fixtures 114 , 116 and 118 may illuminatewall 112 from which light and/or other optical rays are diffusely scattered on movingperson 120 . Thewall 112 may include anopening 122 through which a person may enter and exit the survey field. Sports equipment that may be present may also leave the measurement location through theopening 122 due to the exercise performed. In addition towalls 112, the measurement site may also include a ceiling wall (not shown in FIG. 1), acanopy 124 for varying brightness to affect lighting conditions, and at least onedoor 126 and 128 for closingopening 122 as desired.

图像处理单元110可以给摄像器材100、102、104和106提供时钟信号以便对图像数据定时。于是，不同摄像器材的图像数据可以被同步处理。时钟信号也可以被提供给灯具114、116和118，用来调整由该灯具产生的亮度适用于成像。Image processing unit 110 may provide a clock signal tocamera equipment 100, 102, 104, and 106 for timing the image data. Thus, image data of different camera equipment can be processed synchronously. Clock signals may also be provided toluminaires 114, 116, and 118 to adjust the brightness produced by the luminaires for imaging.

假定要测量高尔夫击球，测量例如可以按照以下方式进行。做运动的人120进入行为地点108。当他或她完成击球时，测量装置可检测运动并自动测量包括运动在内的行为。或者，测量装置可以在行为发生前被自动调到测量状态，随后可以完成该击球运动并且使测量装置在行为发生后脱离测量状态。在测量包括运动在内的行为之前，可以将关于行为人和球杆的预定数据输入测量装置的图像处理单元110中。这些数据可能包括人的体格(如高度和/或四肢长度)、球杆质量(例如长度和/或刚度)、运动器材的质量(如球的结构)。这些数据也可能是要在测量中采用的预定参数。Assuming that a golf shot is to be measured, the measurement can be performed, for example, as follows. The person doing theexercise 120 enters theactivity location 108 . When he or she finishes hitting the ball, the measurement device detects movement and automatically measures behavior including movement. Alternatively, the measurement device may be automatically brought into the measurement state prior to the action, and the stroke may then be completed and the measurement device removed from the measurement state after the action. Before measuring behavior including sports, predetermined data on the behavior person and the club may be input into theimage processing unit 110 of the measurement device. Such data may include a person's physique (eg height and/or limb length), club mass (eg length and/or stiffness), sporting equipment mass (eg ball construction). These data may also be predetermined parameters to be employed in the measurements.

当可以用至少两个摄像器材100、102、104和106拍摄包括运动在内的行为时，图像处理单元110可以确定关于包括运动在内的行为的预定参数值，由此提供测量数据。可以在行为人、球杆和运动器材如球的数十个点上测量这些参数，如果球杆和运动器材在行为动作过程中出现在行为地点108上的话。这些参数例如可以包括击球人和可能有的球杆的不同部分的姿态、速度和/或加速度值。不同部分的最大测量值可以进行相互比较和/或在不同部分上在某个时刻测量的值可以进行相互比较。参数例如可以是：球杆130杆头132在打击球134时的速度，肩膀的最大扭转角度，臀部的最大扭转角度，手臂的最大扭转角度，肩膀的最大速度，臀部的最大速度，手臂的最大速度，肩膀速度、臀部速度和手臂速度和球杆速度的相互关系以及速度变化如加减速度，初始速度，球的初始角度和旋转，等等。可以作为角速度(°/s或rad/s)或线速度(m/s)来测量这些速度。同样，可以作为角加速度(°/s²或rad/s²)或线加速度(m/s²)来测量加速度。这些参数也可以是从其它参数中导出的参数。When the activity including movement can be photographed with at least twocamera apparatuses 100, 102, 104 and 106, theimage processing unit 110 can determine predetermined parameter values regarding the activity including movement, thereby providing measurement data. These parameters can be measured at dozens of points on the performer, club and sporting equipment such as a ball, if the club and sporting equipment are present at theperformance site 108 during the performance of the performance. These parameters may include, for example, attitude, velocity and/or acceleration values of the striker and possibly different parts of the club. The maximum measured values of different sections can be compared with each other and/or the values measured at a certain time on different sections can be compared with each other. Parameters can be, for example: speed of the club 130, club head 132 at impact with theball 134, maximum twist angle of the shoulders, maximum twist angle of the hips, maximum twist angle of the arms, maximum velocity of the shoulders, maximum velocity of the hips, maximum velocity of the arms Speed, shoulder speed, hip speed and arm speed and the relationship between club speed and speed changes such as acceleration and deceleration, initial velocity, initial angle and spin of the ball, etc. These velocities can be measured as angular velocity (°/s or rad/s) or linear velocity (m/s). Likewise, acceleration can be measured as angular acceleration (°/s² or rad/s² ) or linear acceleration (m/s² ). These parameters can also be parameters derived from other parameters.

在图像处理单元中产生了三维的廓影点组，其表示由各摄像器材提供的物体。各人体部分被加入由所述点组构成的图中并且被相互连接起来。加入和相互连接可以通过缩小已知的人体铰接模型与由摄像器材产生的信息之间的差异来完成。In the image processing unit a three-dimensional set of silhouette points is generated, which represents the object provided by the respective camera equipment. The individual body parts are added to the graph formed from the set of points and are connected to each other. Joining and interconnection can be accomplished by bridging the gap between known human articulation models and information generated by camera equipment.

图像处理单元110可以例如按照以下方式确定参数值。例如假定已知高尔夫球的初始速度/方向。于是，可以通过检查图像信息中的瞬时变化在视频帧中搜索运动物体。通过检查许多帧，可以在视频上的运动物体中区分出由每个摄像器材检测的球轨迹。因为同一物体是由经过校准(摄像器材的位置、取向和内部性能是已知的)和同步化(同时检测)的多个摄像器材检测的，所以可以与击球时刻相关地在不同时刻确定球的三维位置。由此可以容易地推导出例如球的初始速度，因为摄像器材的帧速是已知的。可以相应地确定初始角度。在球自由飞行之前，也可以例如通过监视球杆头运动来确定影响球的因素。Theimage processing unit 110 can determine parameter values, for example, in the following manner. For example assume the initial speed/direction of the golf ball is known. Thus, it is possible to search for moving objects in video frames by examining temporal changes in image information. By examining many frames, the ball trajectory detected by each camera rig can be distinguished among moving objects on the video. Because the same object is detected by multiple cameras that are calibrated (the location, orientation and internal properties of the cameras are known) and synchronized (detected simultaneously), the ball can be determined at different times in relation to the moment of impact three-dimensional position. From this, eg the initial velocity of the ball can easily be deduced, since the frame rate of the camera equipment is known. The initial angle can be determined accordingly. Factors affecting the ball may also be determined, for example by monitoring the club head movement, before the ball is in free flight.

为了所述预定参数值，图像处理单元110测量每个行为，利用能适于可视的非监督学习的投影法并依据所述参数分析每个行为。图像处理单元110可以将每个合成测量点加到例如投影法行为映射上，在这里，不同的行为是相互成比例的。所用的行为映射例如可以是神经计算自组织映射、基于sammon映射法形成的映射、GTM(General Topographic Mapping，生成式拓扑映射)、LLE(局部线性嵌入)映射、等距(Isomap)映射或者基于主成分分析的映射。For said predetermined parameter values, theimage processing unit 110 measures each behavior, analyzes each behavior in terms of said parameters using a projection method that can be adapted for unsupervised learning of visualization. Theimage processing unit 110 may add each composite measurement point to, for example, a projective behavior map, where the different behaviors are scaled to each other. The behavioral mapping used can be, for example, neural computing self-organizing mapping, mapping based on sammon mapping, GTM (General Topographic Mapping, generative topology mapping), LLE (local linear embedding) mapping, isometric (Isomap) mapping or based on the main Mapping for compositional analysis.

Sammon映射是非线性投影法，它适用于多维数据的比例化并且可以被用作检测数据模式的工具。Sammon mapping is a nonlinear projection method, which is suitable for scaling multidimensional data and can be used as a tool to detect data patterns.

自组织映射(SOM)可以被用于大型多维数据的分析和可视。自组织映射包括一个或多个维度映射单元如神经元，神经元的相互距离和方向取决于所用的量度。自组织映射不需要监督性教学，但在教学过程中，各神经元变得对各教学数据敏感，从而相似的数据具有在映射的特定区域内形成聚类的趋势。Self-Organizing Maps (SOM) can be used for analysis and visualization of large multidimensional data. A self-organizing map consists of one or more dimensional mapping units such as neurons whose mutual distance and orientation depend on the metric used. Self-organizing maps do not require supervised teaching, but during the teaching process, individual neurons become sensitive to each teaching data, so that similar data have a tendency to form clusters in specific regions of the map.

通常，自组织映射是两维的并且呈矩形，因为它允许人们能在显示器上简单观看到，而且理解起来容易，可是映射也可以例如是三维的或环形的。多维数据理解起来会相当困难，如果不是不可理解的话。Typically, the self-organizing map is two-dimensional and rectangular because it allows simple viewing by a human on a display and is easy to understand, but the map can also be three-dimensional or circular, for example. Multidimensional data can be quite difficult, if not incomprehensible, to understand.

图像处理单元110可以将根据测量数据的测量点加到自组织映射上，并且由于测量结果将被显示出来，所以可以与自组织映射的至少一个神经元有关地显示测量结果。Theimage processing unit 110 may add the measurement points according to the measurement data to the self-organizing map, and since the measurement results will be displayed, may display the measurement results in relation to at least one neuron of the self-organizing map.

图2表示自组织映射。该自组织映射表示与映射神经元相关的连续测量的测量点200、202、204、206和208。每个测量点200、202、204、206和208可以对应于一个神经元，因而这些测量点200、202、204、206和208可以按照神经元精度被标示在映射上。在图2的映射中有两个不同聚类210和212，其中的聚类212包括要在训练中获得的击球的神经元，聚类210包括在训练初始阶段中进行的击球的神经元，测量点200代表第一次完成击球。Figure 2 shows the self-organizing map. The self-organizing map represents successively measured measurement points 200, 202, 204, 206, and 208 associated with mapped neurons. Eachmeasurement point 200, 202, 204, 206, and 208 may correspond to a neuron, and thus thesemeasurement points 200, 202, 204, 206, and 208 may be marked on the map with neuron precision. In the map of Fig. 2 there are twodifferent clusters 210 and 212, whereincluster 212 includes the neurons of the strokes to be obtained in training, andcluster 210 includes the neurons of the strokes performed in the initial stage of training ,measurement point 200 represents the first completed stroke.

于是，第一次击球200的完成不同于理想击球。图像处理单元110可以给球手发送关于击球200的实时反馈信息，该反馈信息表示击球中做得好的特征以及需要改进以获得理想击球的特征。反馈信息例如可以表明在右手侧开球的球手的双手已落至右大腿之前的站立姿势、后摆和开始击球的动作对应于理想击球的特征。从此以后，在击球时刻和击球后，球手双手对右肩的反作用明显不同于理想击球的对应特征。依据反馈信息，球手可以对击球第一阶段感到有把握，而着重于有差距的特征，最终获得理想的完整动作并结束于理想击球聚类212中。Thus, the completion of thefirst shot 200 differs from the ideal shot. Theimage processing unit 110 may send real-time feedback information to the golfer regarding theshot 200, indicating features in the shot that are doing well and those that need to be improved to achieve an ideal shot. Feedback information may indicate, for example, that the stance, backswing, and stroke initiation of a golfer teeing off on the right hand side before the hands have dropped to the right thigh correspond to characteristics of an ideal shot. Henceforth, the reaction of the golfer's hands to the right shoulder at the moment of impact and after the impact is significantly different from the corresponding characteristics of the ideal impact. According to the feedback information, the golfer can feel sure about the first stage of the shot, and focus on the features with gaps, and finally obtain the ideal complete movement and end up in theideal shot cluster 212 .

在高尔夫击球训练中，理想击球通常是比赛条件下的所谓的最适击球，例如是低弧球和右曲球。例如在自组织映射中，由具有某种体格的特定类型的球手以特定类型的球杆完成的这些击球处于同一聚类中，该聚类可以由聚类212表示。作为反馈信息，现场的高尔夫球手接收关于完成击球的特征名单，该名单揭示了例如在聚类210中的完成击球与理想击球之间的特征差异。从完成击球到理想击球的过渡通过重复动作来实现，此时对做成功的特征有把握并改进其它特征。In golf stroke training, the ideal shot is usually the so-called optimum shot under the playing conditions, such as a low arc and a slice. For example, in a self-organizing map, those shots made by a certain type of golfer with a certain build and a certain type of club are in the same cluster, which may be represented bycluster 212 . As feedback information, the golfer at the scene receives a list of features about the shot, which reveals, for example, the difference in features between the shot and the ideal shot incluster 210 . The transition from the completed shot to the ideal shot is achieved through repetition, as the successful traits are mastered and other traits improved.

在一个平面上，在欧氏映射中，两个相互靠近的行为在许多方面是相似的。因此，接近理想行为的特征在行为方式上接近理想行为聚类212。相应地，与理想行为有差距的行为的特征以线性方式接近差距行为聚类210。这种反馈信息能实现达致预定目标的单独训练。On a plane, in a Euclidean map, two objects close to each other behave similarly in many respects. Thus, features that are close to the ideal behavior behave close to theideal behavior cluster 212 . Accordingly, the characteristics of behaviors that are far from the ideal behavior approximate thegap behavior cluster 210 in a linear fashion. This feedback information enables individual training to achieve predetermined goals.

在图像处理单元110的存储器中，可以存储许多关于完成行为和理想行为之间的各种差距的文字说明。每种差距可以与一个口头声明对应起来，该口头声明可以字母数字符号的形式显现在显示器140上，或者以可听声音的形式通过扬声器142来显现。相应地，与理想行为特征相关的文字说明可以存储在图像处理单元110的存储器中。于是，可以督促行为人达到理想特征并有利地改进其它特征。In the memory of theimage processing unit 110, many text descriptions about various gaps between the completed behavior and the ideal behavior can be stored. Each gap can be associated with a verbal statement, which can be presented on thedisplay 140 in the form of alphanumeric symbols, or presented through thespeaker 142 in the form of an audible sound. Correspondingly, text descriptions related to ideal behavior characteristics may be stored in the memory of theimage processing unit 110 . Thus, the actor can be urged to achieve the desired characteristics and advantageously improve other characteristics.

第一次击球200的差距例如可能包括以下事实，即根据常识，肩膀没有扭转足够大的角度以及肩膀相对臀部的扭转关系在测量时刻是不正确的。因此，反馈信息告诉高尔夫球手这种不正确的扭转。呈文本格式的反馈信息可以显示在屏幕上。此外，可以向高尔夫球手显示表示行为地点和行为质量的行为映射。例如，自组织映射使行为同映射上的神经元相称，因而在每次行为后，做出包括运动在内的行为的人可以在映射上实时看到他/她的行为中的哪些是理想的，而哪些是有差距的。Disparities in thefirst shot 200 may include, for example, the fact that, according to common sense, the shoulders are not twisted by a sufficient angle and that the twisting relationship of the shoulders relative to the hips is incorrect at the moment of measurement. Thus, the feedback informs the golfer of this incorrect twist. Feedback information in text format may be displayed on the screen. Additionally, a play map representing the location of the play and the quality of the play can be displayed to the golfer. For example, a self-organizing map scales behaviors to neurons on the map, so after each action, a person performing an action, including movement, can see on the map in real time which of his/her behaviors are ideal , and which ones are different.

也可以在击球动作之间实时监视行为人的击球，例如击球200作为一个神经元，或者通过显示一组几个击球动作如击球200-208的综合特征。就此而言，最多达到几秒的延迟被认为是实时的。在多次击球后，可以显示一排单独击球200-208或者一排多次击球的综合特征，其揭示了反馈信息后的击球改进。做到理想的击球动作可能需要成百课时的上千次重复，在这个过程中，击球200可以通过击球202、204和206发展到将处于由理想击球构成的聚类212中的击球208。It is also possible to monitor the agent's stroke in real time between strokes, such asstroke 200 as a neuron, or by displaying a composite signature of a set of several strokes, such as strokes 200-208. Delays of up to a few seconds are considered real-time for this purpose. After multiple strokes, a row of individual strokes 200-208 or a composite profile of a row of multiple strokes may be displayed, revealing improvement in strokes following feedback information. Achieving the ideal shot may require hundreds of lessons and thousands of repetitions, during which shot 200 can be developed throughshots 202, 204, and 206 intoclusters 212 that will be in the ideal shot. Batting 208.

图3A表示数据库，在该数据库中可以存储行为数据。从各个不同的行为地点300和302，可以通过数据网络304将行为数据传输给数据库306。数据网络304可以是用于传输局部产生的数据的局域网，或者是因特网，通过因特网，可以将来自处于世界不同角落的行为地点300和302的行为数据传输到全球数据库306。于是，数据库306可以存储成百万的行为。数据库306的输入数据可以用作神经网络系统的训练数据，该神经网络系统以非监督形式布置。于是，该数据库的自组织映射构成行为模型，该行为模型变得越来越精确，因为测量数据量在增加。根据所收到的数据，例如可以确定理想行为和/或在某些方面与理想行为有差距的行为。为了工作，数据库306可以包括处理器308和存储器310。Figure 3A shows a database in which behavioral data can be stored. From thevarious activity locations 300 and 302 , activity data can be transmitted to adatabase 306 via adata network 304 . Thedata network 304 can be a local area network for transmitting locally generated data, or the Internet through which activity data fromactivity sites 300 and 302 in different corners of the world can be transmitted to aglobal database 306 . Thus,database 306 may store millions of actions. The input data of thedatabase 306 can be used as training data for the neural network system, which is arranged in an unsupervised fashion. The self-organizing map of this database then constitutes a behavioral model that becomes more and more accurate as the amount of measurement data increases. Based on the data received, ideal behavior and/or behavior that deviates from the ideal behavior in certain respects can be determined, for example. In order to operate,database 306 may includeprocessor 308 andmemory 310 .

所有收集的数据可以被加到自组织映射中，该自组织映射可以作为在各行为地点300和302的基准被恢复。于是，每个行为地点300和302的行为可以以直观的关系被加到其它测量行为上。某个可识别的神经元可以由每个人的每个行为构成，由此一来，所述行为可与这个人自己的至少一个在先行为进行比较，或者这个人的行为可与另一个人的行为同量比较。All collected data can be added to an ad hoc map that can be recovered as a baseline at eachactivity site 300 and 302 . Thus, the behavior of eachbehavior site 300 and 302 can be added to other measured behaviors in an intuitive relationship. An identifiable neuron can consist of each behavior of each person, whereby said behavior can be compared to at least one prior behavior of the person's own, or the behavior of the person can be compared to another person's Behavioral equivalent comparison.

在数据库306中，可以根据不同级别的行为人产生各种不同的自组织映射。因此，例如可以在同一个自组织映射中选择N个高手的最佳行为，其中N是大于2的整数。N例如可以是100。例如在高尔夫球运动的情况下，高手是指如此完成最佳击球的众多球手(即特征组合)，即球(运动器材)的路线满足预定的理想路线，例如就高度、长度和右曲偏离目标直线而言。同样，可以产生例如1000个普通行为人的自组织映射。如果在行为地点有人是超级高手(如顶级高尔夫球员)，他的行为可以被显示在一个自组织映射上，该自组织映射是利用众多高手的结果产生的。相应地，如果在行为地点有普通行为人，则他的行为可以被显示在一个自组织映射上，该自组织映射是用同级别的一群行为人产生的。这就能使一个人被加到与他/她同级别的映射中。因而，相关人士也可以理解要发出的反馈信息，他或她将能更好地努力改善行为。In thedatabase 306, various self-organizing maps can be generated according to different levels of actors. Thus, for example, the best behavior of N masters can be selected in the same self-organizing map, where N is an integer greater than 2. N can be 100, for example. For example, in the case of golf, a master is a number of players (i.e., a combination of characteristics) who perform the best shots in such a way that the course of the ball (sports equipment) satisfies a predetermined ideal course, e.g., in terms of height, length and hook In terms of deviation from the target line. Likewise, an ad hoc map of eg 1000 common actors can be generated. If someone is a super expert (eg a top golfer) at the behavior site, his behavior can be displayed on an adhoc map generated using the results of many masters. Accordingly, if there is a common actor at the action site, his behavior can be displayed on an ad hoc map generated with a group of actors at the same level. This enables a person to be added to the map at the same level as him/her. As a result, the relevant person can also understand the feedback to be sent, and he or she will be better able to try to improve behavior.

也可以将图像处理单元110和数据库306安置在一个行为地点，在各测量点之间没有外延联接。在这种情况下，图像处理单元110和数据库306可以用一台计算机实现，其测量、分析并由测量数据产生自组织映射。无论数据库306在哪里，集中式数据库都能做到训练和教导，无论实际的位置和距离是怎样的。It is also possible to arrange theimage processing unit 110 and thedatabase 306 at one activity site without extensive connections between the individual measurement points. In this case, theimage processing unit 110 and thedatabase 306 can be implemented by a computer that measures, analyzes and generates a self-organizing map from the measurement data. Wherever thedatabase 306 is located, the centralized database enables training and teaching regardless of physical location and distance.

图3B表示作为外延系统的测量装置。体育运动俱乐部322可以具有一个或多个道/场区322，例如高尔夫球场，其具有一个或多个行为地点。在每个行为地点320上的摄像器材100、102、104和106要用于拍摄包括运动在内的行为。摄像器材100、102、104和106所产生的视频图像可以被例如传输给行为地点的图像处理单元110，它可以用作该道、该场或该球场上的服务器。与包括运动在内的行为相关且由图像处理单元110产生的数据可以被显示在显示器140上。Fig. 3B shows a measurement setup as an epitaxy system. Asports club 322 may have one or more courses/fields 322, such as a golf course, with one or more venues. Thecamera equipment 100, 102, 104, and 106 at eachevent location 320 are to be used to capture the event including motion. The video images produced by thecamera equipment 100, 102, 104 and 106 can be transmitted, for example, to theimage processing unit 110 of the venue, which can be used as a server on the track, the field or the field. Data related to activities including motion and generated by theimage processing unit 110 may be displayed on thedisplay 140 .

作为服务器的图像处理单元110包括处理器和存储器，该图像处理单元可以是计算机如PC机，并且可以通过LAN(局域网)或WLAN(无线局域网)与至少一个体育运动俱乐部324的服务器326通讯。体育运动俱乐部324的服务器326可以像数据库306那样对所收发的数据起作用，不管是否在体育运动俱乐部324外有联接，或者对于在体育运动俱乐部外没有联接的小型系统来说，体育运动俱乐部324的服务器326可以明确地用作在图3A中介绍的数据库306。Theimage processing unit 110 as a server includes a processor and memory, and the image processing unit can be a computer such as a PC, and can communicate with theserver 326 of at least onesports club 324 through LAN (local area network) or WLAN (wireless local area network). Theserver 326 of thesports club 324 can act on the data sent and received as thedatabase 306, regardless of whether there is a connection outside thesports club 324, or for a small system without a connection outside the sports club, thesports club 324 Theserver 326 can be explicitly used as thedatabase 306 introduced in FIG. 3A.

不过，体育运动俱乐部324的服务器326可以在体育运动俱乐部324外通过可选防火墙328借助数据网络304进行通讯。可以用作集中式服务器且可以包含从许多体育运动俱乐部搜集来的行为数据的数据库可以通过可选防火墙330并借助数据网络304与每个体育运动俱乐部324的服务器326通讯。由于数据可以从每个体育运动俱乐部324的服务器326被传输给数据库306，所以数据库306可以包含例如覆盖地区、省市、国家、大陆甚至整个世界的行为数据。However, theserver 326 of thesports club 324 may communicate via thedata network 304 outside thesports club 324 through anoptional firewall 328 . A database, which can be used as a centralized server and which can contain behavioral data collected from many sports clubs, can communicate with eachsports club 324server 326 viadata network 304 throughoptional firewall 330 . Since the data can be transmitted from theserver 326 of eachsports club 324 to thedatabase 306, thedatabase 306 can contain, for example, behavioral data covering a region, province, country, continent, or even the entire world.

借助数据网络304并通过可选防火墙336、338，测量装置也可以包含一台或多台计算机332和334，它们可以实时或近似实时地显示包括运动在内的、用摄像器材100、102、104、106拍摄的理想行为。这种显示可以按照与在采用行为地点的显示器140时一样的方式完成。每台计算机332和334可被登入系统，例如由图像处理单元110、体育运动俱乐部324的服务器326或数据库306控制的系统。每个用户或每台计算机具有预定权利来使用所述数据和系统。因此，例如教练员可以将图像处理单元110构造成适用于包括运动在内的某种行为，这能更好地评估行为。为了显示所述行为，每台计算机332和334可以带有与图像处理单元110相同的计算机程序，其控制着显示器140。除了教练员外，管理人员或者对相关运动感兴趣的人也可以观看理想行为。By means of thedata network 304 and throughoptional firewalls 336, 338, the measuring device may also include one ormore computers 332 and 334 which can display, in real time or near real time, the movement of thecamera equipment 100, 102, 104 including movement , 106 ideal behavior for shooting. This display can be done in the same way as when using thedisplay 140 of the venue. Eachcomputer 332 and 334 may be logged into a system, such as a system controlled by theimage processing unit 110 , theserver 326 of thesports club 324 or thedatabase 306 . Each user or each computer has predetermined rights to use the data and the system. Thus, for example, a trainer may configure theimage processing unit 110 to be adapted for a certain behavior including sports, which enables a better evaluation of the behavior. Eachcomputer 332 and 334 may carry the same computer program as theimage processing unit 110 , which controls thedisplay 140 , in order to display the described behavior. In addition to coaches, managers or those interested in related sports can also watch ideal behavior.

例如，数据网络304可以是因特网，防火墙的目的是滤除来自测量装置的干扰信息量。不过，没有必要采用任何一种上述的防火墙。For example, thedata network 304 may be the Internet, and the purpose of the firewall is to filter out disturbing traffic from the measuring devices. However, it is not necessary to employ any of the aforementioned firewalls.

于是，可能包含部件100-106、110、140、306、326、332和334的测量装置可以是如图1所示的局域(小)系统，或者是如图3A和图3B所示的全球(大)系统，或是介于两者间的系统。Thus, the measurement setup, possibly including components 100-106, 110, 140, 306, 326, 332, and 334, could be a local (small) system as shown in Figure 1, or a global system as shown in Figures 3A and 3B. (Large) systems, or systems in between.

图4表示方法流程图。在步骤400中，从至少两个不同的方向拍摄做运动的人100，由此提供关于该行为的图像数据。在步骤402中，从所述行为中测量预定参数值以提供测量数据。在步骤404中，关于在先测量的多个行为，显示依据测量数据的至少一个行为。Figure 4 shows a flow chart of the method. In astep 400, the movingperson 100 is photographed from at least two different directions, thereby providing image data about the action. Instep 402, predetermined parameter values are measured from the behavior to provide measurement data. Instep 404, at least one behavior from the measurement data is displayed with respect to the previously measured plurality of behaviors.

该方法可以通过处理器、存储器和适当的计算机程序来实现。或者，图像处理单元110和数据库306可以至少部分地以硬件结构形式通过单独的逻辑元件或一个或多个专用集成电路(ASIC)来实现。The method can be implemented by means of a processor, memory and suitable computer programs. Alternatively, theimage processing unit 110 and thedatabase 306 may be at least partially implemented in a hardware configuration by separate logic elements or one or more application specific integrated circuits (ASICs).

图4的方法可以以计算机程序产品的形式来实现，该计算机程序产品被安装用于图像处理单元110，该程序产品对用于测量包括运动在内的行为的计算机程序进行编码。计算机程序产品可以存储在计算机程序分发装置上。计算机程序分发装置可以用执行该程序的计算机(图像处理单元)读取。该分发装置可以是任何已知的装置，程序可以借此从制造商/销售商被分发给终端用户。分发装置例如可以是可由图像处理单元读取的介质、程序记录介质、可由图像处理单元读取的存储器、由制造商/销售商配发给终端用户(或者图像处理单元)的数据信号或压缩计算机程序包。The method of Fig. 4 can be implemented in the form of a computer program product installed for theimage processing unit 110, which program product encodes a computer program for measuring behavior including motion. A computer program product may be stored on a computer program distribution device. The computer program distribution device can be read by a computer (image processing unit) that executes the program. The distribution means may be any known means whereby the program can be distributed from the manufacturer/seller to the end users. The distribution device may be, for example, a medium readable by an image processing unit, a program recording medium, a memory readable by an image processing unit, a data signal distributed by a manufacturer/seller to an end user (or an image processing unit), or a compressed computer Package.

尽管以上参照附图中的若干例子描述了本发明，但显然本发明不局限于这些例子，而是可以在后续权利要求书的范围内从许多方面对本发明进行修改。Although the invention has been described above with reference to a few examples in the accompanying drawings, it is clear that the invention is not restricted to these examples but it can be modified in many respects within the scope of the subsequent claims.

Claims (20)

Translated fromChinese

1.一种用于测量包括运动在内的行为的方法，包括：1. A method for measuring behavior including movement, comprising:从至少两个不同的方向对做运动的每个人进行成像，由此提供每个行为的图像数据；imaging each person performing the exercise from at least two different directions, thereby providing image data for each action;为了代表每个运动的预定参数值，测量所述图像数据以提供测量数据；measuring said image data to provide measurement data in order to represent predetermined parameter values for each movement;关于在先测量的多个行为，显示依据所述测量数据的至少一个行为。With respect to previously measured behaviors, at least one behavior dependent on said measurement data is displayed.2.根据权利要求1所述的方法，其特征在于，包括：将每个测量数据的测量点加到用作行为映射的自组织映射中，并且关于代表该自组织映射上的行为的至少另一个预定神经元显示所述测量点。2. The method according to claim 1, comprising: adding the measurement points of each measurement data to a self-organizing map used as a behavior map, and at least another A predetermined neuron displays the measurement point.3.根据权利要求1所述的方法，其特征在于，包括：对做运动的人的成像是在没有将图像分析辅助标记安置在人上的情况下进行的。3. The method of claim 1, comprising: imaging the person performing the exercise without placing image analysis auxiliary markers on the person.4.根据权利要求1所述的方法，其特征在于，包括：将与每个行为相关联的测量数据传输给数据库以便存储。4. The method of claim 1, comprising transmitting measurement data associated with each activity to a database for storage.5.根据权利要求4所述的方法，其特征在于，包括：在存储于该数据库中的测量数据的数据库中产生行为映射，并且将该行为映射传输给每个行为地点以便作为基准。5. The method of claim 4, comprising generating a behavior map in the database of measurement data stored in the database, and transmitting the behavior map to each behavior site for reference.6.根据权利要求5所述的方法，其特征在于，包括：在与每个行为相关联的测量数据的数据库中产生神经元，并且由神经元产生用作行为映射的自组织映射。6. The method of claim 5, comprising generating neurons in the database of measurement data associated with each behavior, and generating a self-organizing map from the neurons for use as the behavior map.7.根据权利要求4所述的方法，其特征在于，包括：利用因特网将与每个行为相关联的测量数据从各测量点传输给数据库。7. The method of claim 4, comprising: using the Internet to transmit the measurement data associated with each activity from the measurement points to the database.8.根据权利要求1所述的方法，其特征在于，包括：关于在先测量的多个行为，至少在一台计算机上附加显示依据所述测量数据的至少一个行为，其中该计算机通过数据网络与所述图像处理单元通讯。8. The method according to claim 1, characterized in that it comprises: with regard to the plurality of previously measured actions, additionally displaying at least one action according to said measurement data on at least one computer, wherein the computer is connected via a data network communicate with the image processing unit.9.根据权利要求8所述的方法，其特征在于，包括：将所述图像处理单元构造成借助计算机而适用于包括运动在内的某个行为。9. A method according to claim 8, comprising configuring the image processing unit to be adapted by means of a computer for an activity including movement.10.根据权利要求1所述的方法，其特征在于，包括：显示对行为的口头意见，该意见依赖于所述测量点同至少另一个行为的关系。10. The method of claim 1, comprising displaying a verbal opinion on a behavior, the opinion being dependent on the relationship of the measurement point to at least one other behavior.11.一种用于测量包括运动在内的行为的测量装置，所述测量装置在每个行为地点具有：11. A measuring device for measuring behavior including movement, said measuring device having at each behavior site:至少两个摄像器材，用于从至少两个不同的方向上对做运动的人进行成像，由此提供关于所述行为的图像数据；at least two camera equipment for imaging the person performing the exercise from at least two different directions, thereby providing image data about said activity;相关的图像处理单元，它被构造成从每个图像数据中测量代表运动的预定参数值以提供测量数据并且关于在先测量的多个行为显示依据所述测量数据的至少一个行为。An associated image processing unit configured to measure a predetermined parameter value representative of motion from each image data to provide measurement data and to display at least one behavior dependent on said measurement data with respect to a plurality of previously measured behaviors.12.根据权利要求11所述的测量装置，其特征在于，所述图像处理单元被构造成将每个测量数据的测量点加到用作行为映射的自组织映射上，在此，所述图像处理单元被构造成使不同的行为相互成比例并且关于代表自组织映射上的行为的至少另一个预定神经元显示所述测量点。12. The measurement device according to claim 11, wherein the image processing unit is configured to add the measurement points of each measurement data to a self-organizing map used as a behavior map, wherein the image The processing unit is configured to scale the different behaviors to one another and to display said measurement point with respect to at least one further predetermined neuron representing the behavior on the self-organizing map.13.根据权利要求11所述的测量装置，其特征在于，所述测量装置被构造成拍摄做运动的人的图像，而不需要将图像分析辅助标记安置在人上。13. The measuring device according to claim 11, characterized in that the measuring device is configured to capture an image of a person exercising without the need for image analysis auxiliary markers to be placed on the person.14.根据权利要求11所述的测量装置，其特征在于，所述测量装置包括数据库，每个行为地点的图像处理单元被构造成将与每个行为相关的测量数据传输给所述数据库以便存储。14. The measurement device according to claim 11, characterized in that the measurement device comprises a database, and the image processing unit of each behavior site is configured to transmit measurement data related to each behavior to the database for storage .15.根据权利要求13所述的测量装置，其特征在于，所述数据库被构造成由存储于所述数据库中的测量数据产生行为映射并且传输所述行为映射至每个行为地点以便作为基准。15. The measurement device according to claim 13, wherein the database is configured to generate a behavior map from measurement data stored in the database and to transmit the behavior map to each behavior site for reference.16.根据权利要求15所述的测量装置，其特征在于，所述数据库被构造成由与每个行为相关的测量数据形成神经元并且由神经元产生用作行为映射的自组织映射。16. The measuring device as claimed in claim 15, characterized in that the database is configured in such a way that the measurement data associated with each behavior form neurons and the neurons generate a self-organizing map as a behavior map.17.根据权利要求13所述的测量装置，其特征在于，每个图像处理单元被构造成用因特网将与每个行为相关的测量数据传输给所述数据库。17. The measuring device according to claim 13, characterized in that each image processing unit is configured to transmit measurement data relating to each activity to the database using the Internet.18.根据权利要求11所述的测量装置，其特征在于，所述测量装置包括至少一台计算机，所述计算机被构造成通过数据网络与所述图像处理单元通讯，以便显示包括运动在内的行为。18. The measuring device according to claim 11, characterized in that the measuring device comprises at least one computer configured to communicate with the image processing unit via a data network in order to display images including motion Behavior.19.根据权利要求18所述的测量装置，其特征在于，所述图像处理单元可被构造成通过所述计算机适用于包括运动在内的行为。19. Measuring device according to claim 18, characterized in that the image processing unit can be adapted by the computer for activities including movements.20.根据权利要求11所述的测量装置，其特征在于，所述图像处理单元被构造成显示对行为的口头意见，所述口头意见依赖于所述测量点同至少另一个行为的关系。20. The measurement device of claim 11, wherein the image processing unit is configured to display a verbal opinion of an action, the verbal opinion being dependent on a relationship of the measurement point to at least one other action.

Images